Beyond images: fetal ultrasound video understanding for detecting congenital heart diseases

Thesis

Congenital heart disease (CHD) is the leading cause of infant mortality among congenital anomalies, yet many cases are missed during routine prenatal ultrasound screening. Routine fetal scans are primarily designed for general anatomical assessment and are not specifically focused on detailed evaluation of the fetal heart. As a result, imaging of fetal cardiac structures is often limited in both coverage and the level of detail required for reliable diagnosis. Detection is further challenged ...

Expand abstract

Authors: Mishra, D

• DOI: 10.5287/ora-m89e68qj0
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: ultrasound video understanding; healthcare video understanding; self supervised video understanding; dual conditioned diffusion models; visual query based video clip retrieval; video understanding; fetal ultrasound video understanding; temporal localization; fetal ultrasound
• Subjects: Deep learning (Machine learning); Medical Video Understanding; Computer vision; Video Understanding; Artificial Intelligence